Do you really need JSON? Sometimes you do: you have complex objects with levels of nesting. Or maybe you have sparse data and JSON is an efficient way to represent it. But sometimes you have a flat data table. A simple spreadsheet. A CSV, if you will. JSON will represent your data, but it will include the field labels for every single record, over and over and over.
JACODA is an answer to that. Maybe not a good answer. A way to represent a flat table in a Javascript-ready format.
See? Exactly the same.
Now look at your brower's console. Easily converted to JSON if needed. Exactly the same.
In theory? Reduced disk space, bandwidth, license usage. All things you pay for. Maybe other benefits too?
Even comparing my small contrived example, the JSON data, with unnecessary whitespace removed, is over twice the size of the JACODA data (header row excluded). That may not make a lot of difference for bandwidth (gzip compression) or disk space (various compression methods), but it will make a big difference for Splunk license usage, for one specific example.
In practice? The differences probably aren't noticeable enough to care. See JACODA versus JSON tests. And we don't need YAJN - Yet Another Javascript Notation.
It was worth exploring, even as a gag.
Inspired by this conversation on Mastodon: What if most JSON data is really just a CSV?